Organizations are exploring the value of adding generative artificial intelligence to their workplace. Questions around reconciling how to do it are the challenge, according to a study from generative AI platform Writer.
The survey asked 1,600 active generative AI users across a range of sectors — financial, retail, health care and technology included — about how adoption is going and what barriers stand in the way. An overwhelming majority of respondents said they were optimistic about their company's plan to adopt AI.
The problem, however, is upper management and on-the-ground employees reported fundamental differences regarding how adoption was going, the quality of the products, and whether a plan to use AI actually existed.
Approximately three-quarters of organizations reported some stumbling blocks with onboarding AI, Writer found. And only 45% of employees thought AI adoption over the past year was going well — compared to 75% of C-suite respondents.
Kevin Chung, the chief strategy officer for Writer, said one of the main drivers of this problem is a difference in understanding of how day-to-day operations might work at a company. Executives want to be seen as knowledgeable about AI, he said, but may not understand how employees in different operations do their work or which tools are most helpful in their work.
"They want on paper for everything to look good, like we're building own on AI systems, we’ve got this AI chat bot out there, it’s working really well," Chung said. "That's what executives are constantly telling their board, or their CEO."
"I think there's like, a will that people want this thing to work," Chung continued, referring to AI adoption. "But then when you pitch it to the people who are using it day to day, they're like, 'No, this is not the tool I want. … I’m just going to buy my own tool and spend my own money on it.'"
These attitudes are in line with some research around AI attitudes in the workplace.
In 2024, Slack's Workforce Lab found only 7% of workers considered AI trustworthy for their tasks and 40% thought their workplace had no AI usage guidelines, despite being optimistic about its potential.
Meanwhile, McKinsey & Company reported high interest in adding AI by employees, but only 30% of AI onboardings succeed, in part due to leadership challenges. The Pew Research Center found more workers are worried than hopeful about AI's use in the workplace, but those who used it saw some productivity gains.
Assessing benefits
Writer's findings indicate businesses are looking to quickly spin up AI programs, with executive-level respondents citing more revenue opportunities, employee interest, competitive pressure and customer demand as reasons to adopt. And there is near-universal belief that AI has created benefits for the company — 88% of employees and 97% of executives saw at least one improvement in their work from using AI.
But the reasons employees are resistant to AI tell the story of why two-thirds of executives said AI has caused division at their workplaces. They listed AI diminishing their value or creativity as a top reason, followed by concerns of job replacement, security issues, poor-quality AI tool and AI adding to their workload.
Chung said the difference in perceived benefits versus employee concerns comes from an inherent tension between those who want to pursue innovation in workplace technology versus the "regulators," or those who make sure the company's structures are protected whenever new technology is introduced.
Innovators "want to use the lates tools, because they know these tools can be very powerful," Chung said. "And then on the flip side, you have structures and organizations who need to think of safety, privacy and unbiased products, because (they) don't want to put something in the hands of their employees or my end users that could potentially compromise companies."
Other disconnects showed up over who holds the power in AI-related decisions and awareness around AI strategy. On general decision-making power, 67% of executives saying it resides with C-suites compared 35% of workers. When it comes to awareness of strategy, 89% executives say they are aware of an existing framework while 57% of employees agreed.
The potential friction points affect companies' AI adoption directly, Writer found.
A third of employees admitted they sabotage their workplace's AI strategy by refusing to use the tools, tampering with the tools' performance or using a nonapproved tool to do company work. Return on investment is also an issue as approximately one-third of executives said they have seen improvements in revenue, employee productivity and cost savings.
'Align on the strategy together'
Chung said this does not point to a need to slow down on AI adoption, but instead a crucial need for organization to get on the same page before doing so. It also means developing good relationships with vendors to really understand what products do and engaging with team members to know what best suits their needs.
"You need to have the IT leaders and the business leaders align on the strategy together, not in silos. And once you're aligned on it, then you should move with speed, but you should not move with speed unless you have that alignment," Chung said. "Because what will inherently happen is if you move with speed but basically have no buy in from the business users, then you're going to, most likely run into a situation where they're going to build something that nobody wants to use."
There is consensus on AI approach as 90% of those surveyed said they are optimistic about their company's AI plans.
"Most people I talk to are like, 'I think we're making the right steps, I feel like we're going in the right direction,'" Chung said. "'We're just not there, yet.'"
You can download and view the whole survey here.
Caitlin Andrews is a staff writer for the IAPP.
